2b:null 2c:["$","title",null,{"children":"Operationalizing AI round table at Booz Allen bit.ly/aiba2020Jan13 | Jonas Almeida | Observable"}] 31:T42e,md` ### Q3 - How does your agency address security, scalability, testability, and monitorability? By a) approaching AI with the serverless execution model (AI as a managed cloud service) and b) developing applications with continous integration/delivery (CI/CD, typically with continuous deployment from github). Failure to address those challenges often drives the funding of Cloud adoption in the first place. In international projects this requires a stateless interoperability model that distributes governance. (provocative) On both accounts things can go badly wrong, particulalry with containarized solutions. It goes without saying that at this point on-prem is completely out of the question. **Note:** in the biomedical arena, the narrative is that of FAIRness. And with that comes issues of privacy during execution (party trick). `32:[] 2f:{"id":"b3bd3dcfc15972f5","slug":"aiba","title":"Operationalizing AI round table at Booz Allen bit.ly/aiba2020Jan13","thumbnail":"c9276a366f34f4a0770c400ae63c14b1c9ebb837726ba718e6edb0d148b79353","access_level":"public","create_time":"$D2020-01-10T20:11:32.609Z","update_time":"$D2020-01-13T15:34:03.567Z","view_time":null,"first_public_time":"$D2020-01-10T20:11:38.066Z","fork":false,"stars":0,"starred":false,"owner":{"type":"team","id":"0f2e596d5bd6ef6e","login":"episphere","name":"Jonas Almeida","avatar_url":"https://avatars-next.observableusercontent.com/3a0ff5abff6da0cdbd100da305ba323e12d43c8fcb7604275e6c72591db9dc69","tier":"starter_2024"},"version":125,"publish_version":125,"role":null,"authored":false,"authors":[{"type":"individual","id":"4355fe13cb71b11d","login":"episphere","name":"Jonas Almeida","avatar_url":"https://avatars-next.observableusercontent.com/3a0ff5abff6da0cdbd100da305ba323e12d43c8fcb7604275e6c72591db9dc69","tier":null}],"members":[{"type":"individual","id":"4355fe13cb71b11d","login":"episphere","name":"Jonas Almeida","avatar_url":"https://avatars-next.observableusercontent.com/3a0ff5abff6da0cdbd100da305ba323e12d43c8fcb7604275e6c72591db9dc69","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"title":"Operationalizing AI round table at Booz Allen bit.ly/aiba2020Jan13","cells":[{"id":0,"value":"md`# Operationalizing AI round table at Booz Allen [bit.ly/aiba2020Jan13](http://bit.ly/aiba2020Jan13)\n## Questions\n\n1. What did you find most challenging when implementing AI solutions?\n2. Who are initial stakeholders to engage as your agency prepares for a successful implementation? \n3. How does your agency address security, scalability, testability, and monitorability?\n\n\n\n\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":10,"value":"md`\n### Q1 - What did you find most challenging when implementing AI solutions?\n\nIt used to be the configuration of AI engines, then data gathering and wrangling. Now it clearly is integrating a governance model that faces the consumer, including researcher, caregiver.\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":20,"value":"md`\n### Q2 - Who are initial stakeholders to engage as your agency prepares for a successful implementation? \n\nThose with a funded mandate towards Cloud Adoption are the initial stakeholders: along with BigData aggregation, AI is typically seen as the natural reward of Cloud adoption.\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":25,"value":"$31","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":105,"value":"md`\n## Round table Notes ##\n* Trustworthy operationalization.\n* AI as enhanced data socialization.\n* Consumer-facing cloud middle-layers where syntactic interoperability and governance (OAUTH scoping) happens.\n* Operationalizing patient \"reported\" signs represent a particular edge in this regard (keynote's \"data exhaust\"). AI as the chief annotator.\n* Public APIs, from NIH All-of-Us to discharge reports for 20M+ patients in NY, census data and outcomes.\n`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W32","stdlib":"1","titling":"auto"}} 30:"$2f:body"